400022 SE Causal inference (2025S)

LECTURER: Daniel Auer
This course introduces PhD students to advanced questions of causality in the social sciences. The course welcomes students of different disciplines within the social sciences. We will critically(!) discuss seminal, provocative, and original empirical studies on key societal questions, such as the causes and consequences of social and political behavior, immigration attitudes, and the (unintended) consequences of public policies. In the morning sessions, we will get to know contemporary quantitative methods that have been introduced to allow for causal statements. First, we establish the experimental ideal and under which conditions the analysis of non-experimental observational data can be interpreted in a causal way. Afterwards, we will discuss specific methods: Randomized Controlled Trials, the difference between (field) experiments and natural experiments, and several quasi-experimental approaches, such as difference-in-differences designs and instrumental variables. During the applied sessions in the afternoon, we will replicate studies (using STATA or R) and investigate if the key assumptions for causal statements are fulfilled. These replication exercises followed by short student presentations and a general discussion will help us identify the challenges we face as researchers when advancing from correlation to causation. The goal of this course is to provide participants with the methodological knowledge and the practical skills to conduct quantitative research and derive causal statements on their own. Participants should have prior knowledge of linear regressions. Textbook chapters will be distributed before the course starts.